Quantifying Cloud Chemistry Processes and Aerosol Optical Properties Using a Particle-Resolved Model
Every cloud droplet contains at least one aerosol nucleus, and the composition and mass of this nucleus can be changed during a cloud's lifetime by several chemical and physical processes. Once the cloud evaporates, a modified aerosol population is released into the atmosphere compared to the population that formed the cloud, which may also have different impacts on climate. For the first part of my thesis, I used a particle-resolved process model, PartMC-MOSAIC, to study the effects of aqueous chemistry and coagulation on aerosols within clouds. PartMC is a Lagrangian parcel model that tracks the evolution of composition and sizes of individual particles and droplets due to emission, coagulation, gas-particle partitioning, water condensation, aqueous phase chemistry and dilution. Results show that due to the formation of ammonium sulfate and nitrate in the cloud, particles that formed cloud droplets grew larger and were more likely to form cloud droplets in future cloud cycles. Overall, cloud processing by aqueous phase chemistry produced aerosol populations that look more homogeneous in terms of composition. Brownian coagulation occurred between the interstitial particles and cloud droplets, but had negligible effects on particle diversity.
The second part of my thesis systematically quantifies the impact of aerosol mixing state on aerosol optical properties. Mixing state describes the distribution of chemical species across a population of particles. Calculations of the aerosol direct effect on climate rely on simulated aerosol fields which can have very different mixing state. However, the model representation of simplified aerosol mixing state potentially introduces large uncertainties into these calculations. To estimate the error in aerosol optical properties introduced by simplified mixing state assumptions, I created a large number of model scenarios using the particle-resolved model PartMC-MOSAIC. To cover a wide range of possible mixing states, I varied 41 input parameters that govern aerosol aging. The combinations of these variables for the individual scenarios were created by Latin Hypercube sampling. Results show that the absorption coefficient can be overestimated by up to 80% when neglecting the diversity of particle composition within the population, and this is mainly due to the redistribution of black carbon. The redistribution of the non-absorbing aerosol leads to the decreasing of scattering ability by up to 30%.
